Emotion and sincerity in Persian discourse: accomplishing ... - CiteSeerX

Emotion and sincerity in Persian discourse: accomplishing the representation of inner states WILLIAM O. BEEMAN

Abstract This paper deals with the presentation and evocation of emotion in performative face-to-face linguistic communication in Iran. Performance in linguistic communication is shown to involve the speaker's need to convey an impression of his or her own inner states. Since a€ectivity is one of the most dicult things to convey in face-to-face interaction, it is posited that a person employing successful linguistic performance skills must have a series of strategies available for demonstrating that he or she is truly conveying a speci®c intended emotion. This involves a two-stage process in which the speaker ®rst signals that a message conveys an emotion, then signals the nature of the emotion being conveyed. In order to accomplish this, culturally prescribed symbolic elements are presented by the speaker that must be performed for others to ``read'' the emotional content of a communication. Added to the performative skills needed by the speaker is the requirement that the emotion conveyed be perceived as ``sincere.'' This paper continues earlier research (Beeman 1986) demonstrating the e€ectiveness of speakers in Iran in creating the contexts for the interpretation of their own strategic communication. Performance, emotion, and linguistics Formalist linguistics is greatly hampered in its ability to handle a€ectivity, because expressions that demonstrate states of human inner feeling are not merely cognitive in nature. They must be performed to be adequately communicated and understood. For this reason, performance is an essential component of emotion in language. A€ectivity is a communicative dimension in language that is by nature systemic. In order for it to be understood, it requires the active participation of both addresser and addressee. In other words, a€ectivity 0165±2516/01/0148±0031 # Walter de Gruyter

Int'l. J. Soc. Lang. 148 (2001), pp. 31±57

32 W.O. Beeman must be performed by an actor, and accomplished in its e€ects on an addressee. Both must establish a common basis for understanding Ð a ``normal'' level for interaction in order to understand what constitutes ``marked'' communication Ð in this case ``marked'' for a€ectivity. This kind of explanation is dicult for much of standard linguistic theory to handle because it is not code-based, nor is it easily subjected to formal analysis. Many of the ``marking'' devices in the indication of a€ective dimensions are nonverbal. Moreover, this kind of explanation requires that the analyst take into consideration the ``state of mind'' of both the addressers and addressees Ð something that has not been easily treated even in pragmatic analysis. For anthropology the diculty lies in the fact that meaning in this kind of communication situation is a cocreation of actors, some of whom may be the ``active'' agents in presenting communicative material, while the other(s) is (are) largely receptive. The communication may be continually modi®ed in the course of its execution by all parties. The result is not a clear and unambiguous reading of easily analyzable symbolic materials, but rather a jostling and jockeying for meaning in which de®nitive understanding may be only an elusive goal. It is natural that a€ective expression should have this quality. After all we can never really know what another person is feeling. Furthermore, it seems to be a feature of human nature that one's true emotions be concealed from others from time to time for personal or for cultural reasons. Even when one has the fervent desire to reveal one's innermost feelings it may not be possible to put them into words or action. For this reason the expression of a€ectivity is a complex performance of culturally habituated interaction routines consisting of the performance of discourse routines of the kind detailed by Deborah Tannen and others (Tannen 1984, 1989): direct statements, metaphorical expression, nonverbal signals, and actions, all subject to modi®cation through registration in communication. Formalist domination of linguistics and sociolinguistics Linguistics emerged as a ``cognitive'' discipline lying somewhere in the vast territory between pure mathematics and experimental physiological psychology at the advent of transformational-generative theory in 1957. The subsequent drive to discover basic rule-governed structures and mathematical/logical principles underlying the production of linguistic forms left little space for discussion of soft, idiosyncratic things like a€ectivity.

Emotion and sincerity in discourse 33 Sociolinguistics has been infected with the problem of formalist domination as well. One case in point concerns Paul Friedrich's brilliant analyses of Russian pronominal usage (Friedrich 1966, 1972). Friedrich's essays on this topic were an attempt to show how pronominal alternation in Russian literary works served to show an enormous range of psychological and cultural dynamics, including status relations, emotional feeling, and rhetorical manipulation. Friedrich's work, inspired by the now classic research by Brown and Gillman (1960) on pronoun usage points out that pronoun usage in Russian incorporates ten dimensions: _ the topic of discourse, the context of the speech event; then age, generation, sex and kinship status; then dialect, group membership and relative jural and political authority; and ®nally, emotional solidarity Ð the sympathy and antipathy between the two speakers (Friedrich 1966: 229).

One of the ®rst reanalyses of this (Ervin-Tripp 1969) immediately reduced Friedrich's rich discussion to a computer ¯ow chart, where the choice between the second-person pronouns vy and ty are reduced to decision nodes in a matrix. Another reference to Friedrich's work by M.A.K. Halliday characterizes it as ``relating the kind and number of kinship terms in general use to changes in the structure of social relationships in Russian society (Halliday 1978: 75). Characterizing Friedrich's analysis as a decision matrix on a ¯ow chart or as a permutation of social relationships does not begin to do justice to the richness of his discussion, particularly in the area of the e€ective use of lexical terms for indicating emotional relations among actors in social situations. Friedrich points out that by studying the dynamic aspects of pronominal usage, much can be learned that is not understood through static models. Two of his observational foci are worth highlighting here. The ®rst has to do with changes in usage over the course of a single interaction: _ the numerous cases of dramatic, rapid and often erratic switching and ``pronominal breakthrough'' often suggest the hierarchical relations between the discriminations. In other words, just as ``regular usage'' symbolized a sucient and necessary co-occurence of discriminations, so the many cases of switching symbolized some realignment, or a change in relative power, or simply the addition or subtraction of a component (Friedrich 1966: 239).

Friedrich's examples of switching are singularly interesting, for most often it is a sudden moment of heightened emotion Ð a reunion, surprise, the realization of love, stressful anxiety, or anger that prompts the

34 W.O. Beeman ``breakthrough'' to the use of the unexpected pronoun: some mood, whim, or mental state could make the speaker play with or altogether ignore the usual rules, depending, of course, on his emotional makeup and social sensitivity (Friedrich 1966: 248). Moreover, he describes another kind of switching, which he terms the latent or mental ty, where the more intimate, familiar pronoun is indicated with paralinguistic attitudes even though vy is being said with the lips. The existence of such phenomena has de®nite implications for ®eldwork in sociolinguistics: one of the fallacies of behavioristic descriptive linguistics and of behavioristic social psychology is that, by a sort of convention, the evidence is arti®cially limited to the overt, actually articulated forms. Such an approach _ is apt to lead to distortion in semantics, because so much of meaning is private and never made explicit during the act of speech. _ some of the most trenchant communication involves the combination of one spoken pronoun with paralinguistic features of body movement and intonation that would normally accompany the covert, unspoken pronoun (Friedrich 1966: 251).

Friedrich's work, its excellence notwithstanding, was based entirely on literary sources, with historical travel accounts providing the background for interpretation. When we are faced with the need to interpret emotional dimensions of language from direct observation, a new set of problems arises. The playwright or novelist helps the reader by selecting those details of background or personal history for his characters that aid the reader in interpretation. A novel with an overabundance of descriptive material is simply not good literature Ð it is tedious, and confusing. Literally everything in the literary work should have signi®cance for the reader within the internal system of the novel. In real life, of course, this is far from the case. The situated actor as well as the ethnographic observer are faced with the problem of reducing the noise of the interaction situation. They must somehow select that which is signi®cant out of the in®nitude of information confronting their senses. Fortunately, man is a social being, and the discovery of signi®cant information in interaction is a social process where everyone helps everyone else. Frameworks and performative accounting procedures The most basic principle in social interaction, well established by phenomenologists and ethnomethodologists, is that basically, every action constitutes an account of itself. Actors as a part of their routine, disattended behavior engage in actions that are able to be ``read'' by others

Emotion and sincerity in discourse 35 as constituting that action. For example, a person who is eating signals others continually that he is eating and not doing something else, like washing windows, or singing a song. To the degree that an individual wants to make sure that others know precisely what they are doing, their routine behavior becomes performative. Then the performative representation of their behavior must be accomplished. Often a person may wish to give the impression that s/he is doing one thing when s/he is in reality doing something else. A prime example is the worker who endeavors to ``look busy'' when a supervisor is passing by. Here too, an account of the action is being given, even though the account may not tally with the ``real behavior'' of the individual. Skill plays a vital role here. Some persons are far better at performing these activity roles than others are. These skilled individuals are thus more adept at communication. Children are naturally good at enjoying themselves and showing it. Adults may ®nd it more dicult to express pleasure in activities and show it. Oce workers need not be trying to fool their superiors to perform in establishing the nature of their activities. A diligent worker may adopt a set of attitudes that signals others that s/he is ``busy'' and doesn't want to be bothered. In the area of providing false accounts, spies and actors must at various times during their work be particularly clever at giving the impression they are doing something other than that which they are really doing. A spy must be able to totally avoid giving the impression that he is observing and noting information. Actors must be able to seem to be drinking, eating, making love, or arguing when they are doing no such thing. Culture provides a great help here, in that it establishes a set of readymade cognitive ``frames'' for interaction that do most of the work for actors who need to make it clear to others exactly what they are doing. This provides the basis for calling even these routine behaviors ``twice behaved'' in Schechner's terms (Schechner 1990). Frames are a complex of physical setting, context, and behavioral sequences that set up expectations as to what behavior is going to occur and how it is to be interpreted. Go€man (1974) has detailed the dynamics of these structures very well. Basically, frames can be seen to exist in time for individuals engaged in interaction. They have a beginning and an ending, and, more importantly, they impose a unique logic upon everything that takes place within their purview. The classic ``frame'' in social science literature is the frame of ``play.'' This was ®rst articulated by Gregory Bateson in 1956 (cf. Bateson 1955, 1956), although the concept was clearly the result of a developing dialogue between a number of scientists concerned with cognitive aspects of animal behavior. Animals are perfectly capable of ``play'' in which they replicate

36 W.O. Beeman actions, such as biting, which normally could cause injury. In play, however, no injury occurs. The message ``this is play'' is paradoxical in Bateson's formulation: Now this phenomenon, play, could only occur if the participant organisms were capable of some degree of meta-communication, i.e., of exchanging signals which would carry the message ``this is play.'' The next step was the examination of the message ``This is play,'' and the realization that this message contains those elements which necessarily generate the paradox of the Russellian or Epimenides type Ð a negative statement containing an implicit negative metastatement. Expanded, the statement ``This is play'' looks something like this: ``These actions in which we now engage do not denote what those actions for which they stand would denote'' (Bateson 1955: 40±41).

The a€ective dimension of language is enacted as ``framed'' communicative behavior, just as ``play'' is ``framed'' behavior. In particular when emotions are being expressed, there must be a tremendous reduction in the possible interpretations that can be established by interaction participants, to insure the success of the communication. When a€ectivity is unsuccessfully conveyed, the reason is usually that the cognitive frame for the interpretation of the behavior was insuciently established. Occasionally, emotional expression invades a frame for communication where it is inappropriate. This is an example of what Go€man calls ``¯ooding out,'' and it usually represents a great risk for participants. Usually individuals keep control, however. Playboy magazine's advice column, counting on the general tendency of humans to stay in frame, once counseled a young man who wanted to break o€ a long-standing relationship to give his girl friend the news at lunch in the fanciest restaurant in town. ``She won't dare make a scene,'' the magazine wrote, ``and you'll walk out of the restaurant a free man.'' For anthropologists, the diculty in dealing with a€ective dimensions of language behavior in foreign cultures is particularly acute. Anthropology traditionally eschews value judgments and statements about happiness, anger, jealousy, and a€ection, as components of observed behavior are often seen as at least partly the projections of the observer onto the situation. Nevertheless, people everywhere are successful in conveying their feelings toward others, and the framed understandings that allow the linguistic expression of a€ectivity to be interpreted by others is theoretically accessible to the anthropologist. Jules Henry (1936) in an early essay on the expression of emotion in language indicates that there are some languages that conveniently provide grammatical marking for

Emotion and sincerity in discourse 37 emotional states in speaking. This is also true of Japanese, where a rich variety of linguistic expressions aid speakers in indicating to others exactly what their mental attitude is toward the statements they are producing (while not paying much attention to empirical distinctions, such as those between singular and plural, person and number, etc.) (cf. Seward 1971; Passin 1980; Miller 1967, 1977). Basic frames in Iranian interaction Persian is in some ways like Japanese, in that it is a language with a very simple grammatical structure, and a rich set of idiomatic expressions that help individuals to convey accounts of their feelings. These accounts, as with all accounts, may be correct or incorrect re¯ections of actual emotions. It is a function of all parties in interaction to come up with the correct interpretations for what is said. Persian society (like all societies everywhere) provides for basic frames that clue individuals as to appropriate language behavior for any given situation. These frames provide a cognitive map that helps de®ne what is normal and expected. Departures from normalcy are duly noted and, because of their unexpected nature, carry special signi®cance. The point can be easily illustrated in American society by considering the function of winking. For most Americans a wink indicates collusion, or a special relationship. It has sexual connotations as well and can be considered as a ``pass'' or a come-on. In a singles bar late in the evening, winks don't mean much, because they are expected behavior. In court, at a formal reception, or at a church ceremony, winking is not expected behavior. Therefore, if it is engaged in, it is immediately noticed and carries heavy signi®cance. It is so out of place at these times that it may actually be ignored or thought a mistake or an accident. A€ective language has this quality. Words or expressions that are quite normal in some frames, to the point where no one thinks much about them, acquire heightened signi®cance when they are used in other frames. It is these heightened forms of expression that seem best to convey emotion. It should be added here that silence is likewise an expressive form, and silence used where it is not expected is generally a very e€ective a€ective communicative device. The basic dimensions of Iranian society are not terribly complex in a structural sense, but they provide for a rich play of linguistic expression. The basic orientation of interaction frames consists of a continuum, with situations that are considered more inside at one pole, and situations that are considered more outside at the other.

38 W.O. Beeman The contrast between inside and outside is pervasive in Iranian thinking and governs many other aspects of national life.1 The outside/inside distinction is indeed a common cultural trope in other areas of the world (Indonesia, Japan). However, its particular realization in Iran is unique to Iranian culture Ð particularly in the manner in which it clusters with emotional life, art, religious philosophy, and social demeanor. In philosophical terms, the inside is often referred to as the baten and is revealed as the seat of the strongest personal feelings. The inner peace and joy of the enlighted religious mystic, the Su®, contrasts with the inner turmoil and con¯ict felt to exist in the baten of most individuals. The ®res of romantic passion, indignation, and righteous anger all have their seat in the baten. Still it is to the baten that one must turn to ®nd peace. Despite the seeming contrast between the passions of the baten, the emotions contained there are generally positively valued, and their expression on appropriate occasions is not only socially sanctioned, it is required on occasion. The baten is also the center of social peace in its theoretical architectural form, the ñndñrun. This is the space within the household that is the most private, the most secluded. It is the seat of family intimacy and is where the women of the family may ®nd safety from the outside world. It is thus the place where one can give rein to free expression. The outside is by contrast the realm of controlled expression. It is the realm of politesse, and of proper expression and behavior, where one's true feelings must be controlled, where a proper public face must be put on one's words. The philosophical realm of the outside is labeled the zaher, and it is considered a necessary concomitant of life. Nevertheless it is not highly valued in moral terms. The zaher is the realm of corruption, and of worldly in¯uences, but it is also a bu€er for the delicate world of the baten. Thus one may not value the zaher, but one must know how to operate in it. The admonition zaher-ra hefz kon! `protect external appearances!' is widely given and widely heeded, since by maintaining the external aspects of one's behavior, one can remain out of danger and protect and control the baten. The zaher has its physical re¯ex in architecture as well. It is the birun or biruni, the public reception areas of the household, where strangers may be entertained without endangering the private space of the family. The ñndñrun and birun of the household are to a degree portable. The Tehrani family going on a picnic takes its ñndñrun with it to the outdoors by bringing carpets, cooking utensils, a samovar, and the accouterments of home. A person traveling to another town will have innumerable locations available to him where he can be in an ñndñrun Ð where he can put on pajamas, nap, and eat around a family dinner cloth (sofreh).

Emotion and sincerity in discourse 39 This is usually in the home of relatives or close friends. Thus one of the social considerations of marriage is always the assurance that the members of one family (down to distant cousins) can indeed become admittees to the ñndñrun of the other family. The feeling of being in an inside/ñndñrun/baten situation vs. an outside/ birun/zaher situation is a matter of subjective judgment, depending on a calculus of factors, much as those identi®ed by Friedrich with regard to Russian pronominal usage above. One judges a particular situation based on location, the company one ®nds oneself with, the topic of conversation, and so forth down the list. Moreover, as was stated at the beginning of this section, the two situations represent poles on a continuum. An individual feels himself to be situated somewhere between the extremes, neither totally ``inside'' nor totally ``outside'' for any given instance of interaction. This basic orientation forms the social canvas against which a€ectivity in language is expressed. Iranian personal orientation The expression of emotion, as mentioned above, requires interaction between one or more individuals. This interaction takes place within a framework, but the interaction itself contains the expression of a€ectivity. Just as every action is an account of itself, every communication can be seen as imposing a commitment to a state of a€airs on the part of participants. Bateson likewise pioneered this approach in which every communication can be seen as having a ``report'' and a ``command'' aspect respectively (cf. Ruesch and Bateson 1951: 179±181). Watzlawick et al. summarize this di€erence usefully: The report aspect of a message conveys information and is, therefore, synonymous in human communication with the content of the message. It may be about anything that is communicable regardless of whether the particular information is true or false, valid, invalid, or undecidable. The command aspect, on the other hand, refers to what sort of a message it is to be taken as, and, therefore, ultimately to the relationship between the communicants (Watzlawick et al. 1967: 51±52).

Persian contains a number of stylistic devices that automatically help individuals to signal each other concerning many aspects of their assessment of their relationships to each other. These stylistic devices principally deal with contrasts between relationships that are statusdi€erentiated and relationships that are status-undi€erentiated. I have

40 W.O. Beeman described the status system in Persian and its linguistic re¯exes at length elsewhere (cf. Beeman 1976a, 1982, 1983, 1986) and so will limit my discussion at this point to a brief summary. Hierarchical di€erentiation seems to be a nearly universal feature of human life, but in some societies, hierarchy takes on a special symbolic signi®cance, such as in India (Dumont 1970) or Japan (Lebra 1976). There are few societies that take the obligations of status as seriously as Iranian society. Persons placed in a position of superiority should ideally rise to that position and retain it by ful®lling obligations toward inferiors that insure their support and respect. Inferiors in turn retain their ties to speci®c individuals in superior positions by reciprocal observance of obligations of their own. Social behavior between superior and inferior tends to revolve around patterns of mutual exchange, which can be represented in schematic form as consisting of 1. providing action for the other person; 2. providing material goods for the other person; 3. stimulating others to provide either goods or action. When two parties enter an interaction that de®nes both as status unequals, the exchange becomes di€erentially marked with di€erent meanings attached to each move in the exchange. The person in the superior role engages in 1. providing favors; 2. providing rewards; 3. stimulating the other person to provide goods or action through issuing orders. The person de®ned in the inferior position may exchange goods or service that are entirely equivalent, but they are de®ned di€erently, as 1. providing service; 2. providing tribute; 3. stimulating the other person to provide goods or action through making petitions. Relationships of equality likewise involve exchange in the same manner; the di€erence is that in equality relationships the exchange of goods is non±status-marked, and is absolute. The ideal situation is one where two individuals involved in an intimate relationship anticipate each other's needs and provide all for the other without thought of self. There should be no need for petition or order; and no thought of service, tribute, favor, or reward. Relationships of both equality and inequality can be deeply satisfying in Iranian life. Moreover, obligations in both kinds of relationship are absolute, the ultimate ful®llment coming from a willingness to enter into total self-sacri®ce in meeting the needs of the other person. This is true

Emotion and sincerity in discourse 41 of the superior, who must care for those who look up to him, even if he is ruined as a result; the inferior, who follows his leader in all respects even to his death; and the comrade, who gives his all for the sake of his intimate companion. Like the frame orientations discussed in the last section, these personal relationships tend to represent polar orientations. Few people can ful®ll the absolute obligations of social relationships as described above, though they do represent cultural ideals. They also represent orientations for the kind of linguistic signals that participants in interaction need to inform each other of their assessment of their relationships to each other Ð the ``command'' functions of communication as cited by Watzlawick et al. above. In Persian such signaling is carried out largely through stylistic variation in language. This stylistic variation includes the pronoun alternation analyzed by Friedrich, but much else as well. The two can be seen to interrelate in Figure 1. Figure 1 expresses the tensions speakers feel between their inner feelings and the means they will use to exhibit them in outward expression. The speaker's assessment of the interaction frame ranging from more zaher to more baten ultimately determines the form of that outward expression, but both factors are determinant. Style and emotion Pronouns and verbs in Persian are oriented in three directions, which correspond to the basic orientations in social relations. The ®rst orientation re¯ects relationships of inequality and involves a process of ``other-raising'' vs. ``self-lowering.'' Basically, one uses terms that serve to place oneself in an inferior status, and the other person in a superior status. These consist of a series of substitutions for neutral verbs and pronouns. Thus self-reference may use the expression bñndeh `slave' in place of the neutral pronoun man `I'. Reference to the other person in interaction may substitute the verb fúrmudñn `command' for the neutral verb goftñn `say'. "

inner feeling + ,

#

outward expression interaction frame zaher .ÐÐÐÐÐ- baten Figure 1. Expressive congruities in Iran

42 W.O. Beeman Relationships of equality use parallel terms. Both parties will use the same pronouns and verbs, and these tend to be rougher and less re®ned as intimacy between the parties increases. Once again, this phenomenon has been well described elsewhere, and I leave readers who are interested in the linguistic details to these sources.2 The important aspect of these linguistic stylistic materials is that they are used di€erentially, depending on the context in which interaction takes place. It is thus the interplay of cognitive context, in terms of general orientation to inside vs. outside dimensions, with human orientations toward status and equality that creates the interactional grammar for the expression of emotions in Iranian life. Words in play In general the two frames for interaction, outside vs. inside, articulate with the ``command'' functions of language to introduce a de®nite bias into Iranian interaction. As individuals ®nd themselves in situations oriented more toward the outside/birun/zaher, they are more disposed to signal others that they are in a relationship of inequality. Moreover, expression becomes altogether more restricted in these situations. As individuals move into inside/ñndñrun/baten situations, they are more disposed to signaling that they are in relationships of equality, and expression becomes much freer. These interrelationships form a basic schema for normal interaction, as represented in Figure 2. When interaction is seen to operate according to this schema, it is perceived as operating normally. In linguistic terms, communication conforming to these con®nes is ``unmarked,'' as described by Joseph Greenberg in his in¯uential work, Language Universals (1966). In general, expression of emotion falls within the areas that are outside of the basic interaction schema. Because the intent of the speaker is to communicate a message that is charged with special meaning, he must cast that message in such a way that it will constitute ``marked'' behavior. In order to do this, he has two strategies open to him in communicational terms. INTERACTION FRAME ORIENTATION inside/ñndñrun/baten .ÐÐÐÐÐÐÐÐÐ- outside/birun/zaher status equality status inequality free expression restricted expression Figure 2. Iranian basic interaction schema

Emotion and sincerity in discourse 43 The ®rst strategy consists of violating the basic schema, by presenting linguistic behavior that is appropriate to the polar direction opposite to the one in which the interaction is actually taking place. Thus linguistic behavior appropriate to ``inside'' situations, if employed in true ``outside'' contexts, will convey a€ect. The reverse is also true. ``Outside'' language used in ``inside'' situations will also be read as a message carrying a€ective overtones, as will be demonstrated below. The other strategy involves intensi®cation of appropriate communication behavior. Thus language that overstresses social inequality in outside situations will be read as conveying a€ect. Similarly, language that overstresses intimate equality in inside situations will also convey a€ect. Such strategies may be universal in nature. Certainly they seem to account for a good deal of the pronoun switching cited by Friedrich on brief inspection, but con®rmation of this would depend on development of a complete theory of Russian interaction contexts and interpersonal orientation. Emotion and language Before proceeding with a more detailed analysis of the schema cited in the last section, I think it advisable to address a few comments to the question the di€erences between a€ectivity as felt by the individual, and a€ectivity as expressed in communication. There is a de®nite feeling among some students engaged in the crosscultural study of personality that people in di€erent cultures entertain constellations of emotions unique to their culture. I do not presume to take a de®nite position on this question, but I have argued elsewhere (Beeman 1976b) that ®eldworkers cannot observe emotions or psychological characteristics of individuals directly. One can only observe overt expressions of those more private phenomena. These expressions are organized and regularized in their own way within broad systems of communication, which, in great part, obliterate vast di€erences in individual emotional expression (thus insuring mutual intelligibility, however). Observers can ascertain that an expression of emotion is being carried out, but it may be impossible, or nearly so, to ascertain the precise a€ective content of that expression except in very broad ways. Cross-cultural psychiatric studies are notoriously vague for just this reason. Danish psychiatrist Erling Jacobsen sums up the dilemma as follows: _ we learn to associate feelings with emotionally tainted words in an erratic way. Furthermore, we may be prevented from correcting our private misunderstandings

44 W.O. Beeman and from reaching agreement with others concerning the use of the words, because we cannot transcend the knowledge we get from our own private and more or less impoverished collection of bundles of feelings. In these circumstances, it is small wonder that the behaviorists feel that emotions cannot be the objects of a science, or that Freud preferred to base his metapsychology on the concept of drive, rather than on a theory of emotions (Jacobsen 1979: 218).

The problem is far greater for the observational social sciences. A ®eldworker is continually endangered by his tendency to apply the rules for expression of emotion from his own communication system to the society under observation and thus come up with judgments of the sort identi®ed in the literature as ``national character'' or ``modal personality.'' These are at best relativistic judgments expressing the range of distance that exists between the expressive rules of the culture of the observer and that of the observed. I am hampered in my own description of emotional expression in Iranian society by lack of terms suciently neutral to avoid the overtones that adhere to English words used for expressing emotions. A€ection, anger, sadness, disappointment, etc., are all words that carry a cultural load, but they are all we have at present. In ethnoscienti®c terms (which I am not particularly enamored of, but which suit here), there is no ``etic grid'' for the description of emotions. Therefore, in discussing the Iranian expression of emotional terms, I am reduced to trying to formulate a description that emphasizes the observable e€ects of concrete communications on others, and the ways these are interpreted by individuals involved in actual interaction, in terms of the logic of the representation system itself.

Patterns of Iranian emotional expression In Iranian life, there are orientations that are positively valued, and others that are negatively valued. The Iranian basic interaction schema, characterized above in Figure 2, represents stability within the system of social life. It is a system of high pattern congruity. Thus conformity to this basic schema in interaction represents predictability for interaction participants, and in predictability in Iran, or any culture, lies the security of knowing what will happen, and what one must perform in order to lead a comfortable and secure existence. This means that individuals in outside/birun/zaher situations will feel comfortable using modes of expression that emphasize status inequality and are restricted and overdetermined. Likewise, individuals

Emotion and sincerity in discourse 45 in inside/ñndñrun/baten situations will feel comfortable utilizing expression that emphasizes status equality and is free and open in nature. However, within the basic schema there is a de®nite built-in bias. As mentioned above, Iranians value inside/ñndñrun/baten situations and orientations far more than outside/birun/zaher situations. This is con®rmed again and again in poetry, proverbs, religious philosophy, and humble moralizing (cf. Bateson et al. 1977). As mentioned above, emotional expression is encoded in Iranian behavior in expressions that fall outside of the basic schema, either by transposition of expected behavior, or by intensi®cation of it. Because of the positive bias accorded inside/ñndñrun/baten orientations, expressions outside of the basic schema drawn from that orientation tend to be valued more positively than those drawn from the outside/birun/zaher repetoire. The following cases drawn from ®eldwork in Iran will serve to illustrate: Case 1 A man is arguing with an army ocial over a petition for a military service waiver for his son. The discussion proceeds, the man making himself more and more abject in his petition, to no avail. Finally he breaks down in tears, and begins to address the ocer in familiar terms. The ocer's mood immediately changes, and he quiets the man, telling him he will do what he can for the boy. When asked why he changed his mind, the ocer said, ``Well, it was obvious he was sincere; he was speaking from the bottom of his heart.'' In this case, the petitioner, in an outside/birun/zaher situation, conformed to the basic interaction schema for a long while, and then ``¯ooded out'' into inside/ñndñrun/baten behavior. His message was then clearly read as prompted by sincere emotion and was valued positively. Case 2 Two oce workers are involved in a heated but polite discussion. One makes a disparaging remark toward the other. The second man's reaction is immediate. He rises, and suddenly lunges for the ®rst man. He begins shouting and using foul language and must be restrained by the others. The angry man is quieted and persuaded to make up the quarrel. The two men kiss on the cheek, and calm is restored. Later a spectator, queried about the quarrel, replied, ``He had to behave like that Ð his honor [qeirñt] was o€ended, and that was the only proper way to show it.'' When the man who was attacked was asked about his feelings toward the angry man, he said, ``Oh, now we will be closer than ever.''

46 W.O. Beeman In this case, justi®able anger had to be expressed through violation of the expected behavior in this relatively outside/birun/zaher situation. The situation of ``making up'' after a quarrel is, like the angry outburst itself, behavior drawn from the inside/ñndñrun/baten repetoire. The two parties are indeed likely to be on closer terms after the ®ght since they have just experienced a moment of intimacy that might not otherwise have taken place in their relationship. The introduction of outside/birun/zaher behavior into situations de®ned as inside/ñndñrun/baten is likewise a signal of emotion, most often read as anger or displeasure. Case 3 A mother is angry at her son for arriving home late, therefore she addresses him thus: Xeili bebñxshid, nñfñhmidñm, jenab-e ali kei tñshrif avardñnd. ñl'an sham-ra meyl mifñrma'id? (roughly) `Oh, excuse me, I didn't notice, has his honor favored us with his presence? Will you now condescend to dine?' The son replies, Xub. Mñno bebñxsÏ . Dir kñrdñm. Hñnuz ghñza hñst? (roughly) `OK. Sorry I'm late, is there any food left?' This kind of sarcasm is not unknown even in the West, but it is important to note its e€ectiveness as a conveyor of displeasure. The mother doesn't actually upbraid her son for being late, but she elicits an apology from him none the less, however perfunctory. In many traditional families, the wife, mother, or sister is not supposed to question or express disapproval of the actions of men. The linguistic device of employing outside/birun/zaher speech forms in the intimate setting of the family is far more e€ective than remonstration, which will usually trigger an angry response, even physical violence. Anger in the family is an example of an intensi®cation of internal/ ñndñrun/baten behavior, which the community will largely view as justi®ed on the part of the man. Other examples of intensi®cation in these settings include open protestations of devotion, physical intimacy, and body contact, even, actually one should say especially, between members of the same sex, and the sharing of private experiences such as listening to music, drinking, smoking opium, and reading poetry. These latter activities are clearly associated with experience of the world of the baten and are thus usually carried on only in intimate circumstances with intimate companions, contrasting sharply with the West, where such activities can easily be enjoyed even in the company of relative strangers.

Emotion and sincerity in discourse 47 Such intimate events constitute high points in the life of the individual and are often accompanied by an overwhelming wave of emotional feeling. Weeping is not uncommon in such circumstances, as well as an intense feeling of joy. ``I live for the times I spend with my friends, and at such times, I know without a doubt that I would sacri®ce myself immediately for any one of them,'' claimed one otherwise staid government ocial in an unguarded moment. Intensi®cation of outside/zaher/birun communication is generally used as a device to achieve social control within the society. In this function such intensi®cation is labeled by Iranians as a variety of the polite communication known generally as tñ'arof. Tñ'arof itself is an enormously complex subject, which I have dealt with in other publications (cf. Beeman 1976b, 1982, 1983; see also Hillman 1981). In general, tñ'arof seems to have an aspect that is positively valued, and a side that is negatively valued. In both cases, however, I would maintain that it represents a form of outside/birun/zaher behavior that is more intense than normally expected, and this intensi®ed communication serves to control the behavior of those addressed. Positively valued tñ'arof is often seen as consisting of exercising proper behavior toward others, particularly guests, in using polite language, giving food or gifts; or in paying compliments or showing regard for those who are truly worthy of respect or deserving of such behavior. This variety of tñ'arof is valued because it is viewed as an expression of sel¯essness and humility.3 Nevertheless, such tñ'arof is de®nitely a control device, for it serves to immobilize the recipient. The guest is not allowed to do anything for himself, he is literally at the mercy of his host. He cannot show displeasure, or anger. He must make strenuous e€orts just to leave the scene where he is being entertained. The high-status person is often prevented from learning about many things because he is ``protected'' by those around him. This was de®nitely the case with the deposed shah, Mohammad Reza Pahlavi, who was simply never told about much of the turmoil surrounding his rule, because his underlings only wanted him to hear ¯attering things about himself. Negatively valued tñ'arof may consist of exactly the same intensi®ed behavior, but it is done for manipulative purposes, in order to make incumbent on the other party in interaction the ethics of the superior party in a relationship. Thus a man may be ¯attered into doing another person's will in order to show his ``generosity,'' or ``pity'' for the ``poorest and most abject of his servants,'' to use some of the phraseology. The fact is that the strategy works often enough that it is widely used. Indeed, a skillful manipulator of the language is often extremely e€ective

48 W.O. Beeman in controlling others in just this manner. This use of communication is valued negatively because it is clearly carried out for personal gain. It is in no way an expression of sel¯essness and humility; it is in fact seen as a disguised expression of arrogance. The diculty in identifying this negative aspect of tñ'arof arises from the fact that whatever its intent, if used skillfully, it is very dicult to counter. One cannot say to another, ``No, I'm not the great, magnanamous person you paint me, I am just a poor boob like you!'' For, as implied, one ends up by insulting the addresser. What is most often done in the case where undue ¯attery is being used is to reciprocate to at least the level of the addresser. This initiates a kind of competition that I have described elsewhere as ``getting the lower hand'' (see Beeman l976a; also see Irvine 1973 for a similar process in Wolof ). Intensifying outside/birun/zaher behavior is also used to control individuals who are misbehaving, or acting in a way that is displeasing to the speaker. Thus one can gauge the unhappiness of another party with the course of an interaction by noting this intensi®cation, as his or her behavior becomes more and more cold, distant, and formal. Control of a superior person who is himself expressing anger can be dealt with through this kind of intensi®ed language. By continuing to assert the angry person's superiority and one's own inferiority, the angry person may be able to be placated. In this way, intensi®cation of outside/birun/zaher behavior is a device that gives a double message. It purports to be signalling that the individual using it is expressing an emotional message: a sincere, genuine regard for the status of the other person. This is true often enough to make it good form not to doubt the true intentions of the person exhibiting this behavior. It is often the case, however, that this kind of intensi®cation is used not to express one's own emotions, but to control the emotional state of others. This is accomplished by lulling individuals into serene complacency, by stirring their sense of pity and duty, by warning them that they are getting out of line, or by soothing anger that might spin out of control and do damage to those around them. As with the case of Russian pronouns, both transposition and intensi®cation can be exhibited in communication dimensions other than speech, although Iranian interaction normally prescribes that bodily attitudes be coordinated with speech. As one moves into more outside/birun/zaher situations, as has been mentioned, movement is restricted. The body is sti€, hands are held together in front of the body, and eye contact is avoided. In the Qajar court, ordinary courtiers had to sit on their heels with their feet tucked

Emotion and sincerity in discourse 49 under them in the presence of the shah. The right to sit cross-legged was a royal prerogative (cf. Peterson 1981: 384). Inside/ñndñrun/baten situations likewise involve a loosening of body attitudes. Siblings or cousins, who come closest to being structural intimate equals in Iranian society, literally sprawl on the ¯oor when alone in their own households. Should an older relative enter, an uncle or grandfather for example, the younger members will pull themselves into a more respectful sitting position. In outside/birun/zaher situations even siblings, cousins, and sweethearts will behave toward each other in a more restricted way. They may address each other with formal pronouns or defer to each other in a rather formal way. Still, expressions of intimacy are possible even in formal settings through the use of bodily expressions and vocal dynamics drawn from the repertoire of intimate behavior. Performing sincerity Sincerity is not itself an a€ective state. It nevertheless must be performed in order to be e€ectively accomplished as a representation of behavior. It is rather the assessment by addressees that the expressions of addressers are true representations of their feelings and emotional state. Insincerity is by contrast an assessment by addressees that there is a discrepancy between addressers' expression and their true feelings. Sincerity is thus a paradoxical a€ective expression in that its presence or absence is ultimately ascertained not by the expressor(s), but by the persons to whom it is expressed. Thus it has a performative dimension to it. Very little research has been carried out on this topic (but see Brennis 1988; Irvine 1982, 1990). Some additional paradoxes exist with sincerity. One of the principal ones is that some communication roles, such as that of actor, require that the audience be convinced of sincerity of expression without necessarily believing that the actor is re¯ecting his or her personal emotional state. For some actors schooled in the ``method,'' the ability to call up emotion and then express it on stage is a highly valued skill. Other actors whose work follows more ``technical'' lines are of the opinion that experiencing the actual emotions one is expressing is not necessary; a well-trained actor can be convincing without this added step in preparation.4 Irony is a form of insincerity that is meant to be recognized as such. The addresser demonstrably exhibits behavior that is incongruent with his or her feelings in order to express displeasure with the addressee or with a situation to which he or she wishes to draw the addressee's attention.

50 W.O. Beeman Sincerity seems to be a cross-cultural variable. In many societies sincerity is highly valued for all people at all times. It is expected that a€ectivity and its outward expression will normally be congruent. In other societies, people expect some discontinuity between emotion and expression and may devote a good deal of social energy to sorting out the sincere from the insincere. Realizing that sincerity and its assessment vary widely helps us to understand that there is more going on in an assessment or discussion of sincerity than just inner emotional states and assessments of the congruity between these states and external expression. Power and status relations between addressers and addressees are of paramount importance, as are settings in which emotional display takes place. As White suggests in a recent paper, _ emotions are not simply expressed in social situations, but are constituted by the types of activities and relations in which they are enacted (1990: 64).

Since the assessment of sincerity lies with the addressee, the burden of proof lies with the addresser. He or she risks being judged insincere if his or her demeanor is inappropriate to his or her social role in a given situation. Insincerity is not the only possible judgment of inappropriate demeanor, however. The person exhibiting such behavior might also be thought to be rude, distraught, drunk, insane, or purposefully deceptive. These readings of inappropriate demeanor are telling. The proper response to such judgments on the part of an addressee is for the addresser to reassert and try to establish sincerity. The way he or she does this will be through culturally habituated discourse routines. These reassertions fall under the general rubric of ``repair routines'' in discourse analysis (cf. Tannen 1984, 1989). I wish to explore just such routines in Persian below.

Persian glosses for sincerity There is no word in Persian that exactly translates the English word ``sincerity,'' though I maintain the concept to be universally understood. The closest concepts refer speci®cally to acts of speaking, and hence of discourse: rast goftñn, jeddi goftñn, rast-gu budñn, ñz tñh-e del goftñn. As will be seen below, speakers use these phrases as formulae for asserting sincerity whenever it is questioned or doubted. Although these phrases in one way or another refer to ``truth,'' they go far beyond reference to

Emotion and sincerity in discourse 51 mere assertions of the truth Ð they also refer to personal qualities of the addresser. Insincerity is harder to gloss in Persian. It is usually expressed through descriptions of speci®c acts that demonstrate speci®c insincere acts: do-ru budñn, chaplusi kñrdñn, masmali kñrdñn, and dorugh goftñn.5 This ®nal term is extremely potent in Persian culture. Although it is often glossed as ``lying,'' it is much more serious. To accuse someone directly of durugh goftñn will call forth vehement, angry denials and assertions of sincerity.6 Both sincerity and insincerity are deeply intertwined with the twin concepts of zaher `outside' and baten `inside' treated above.

Performance practice and sincerity Persian speakers have a number of available discourse strategies that they can use to establish both emotional expression and sincerity. If emotion is registered and seemingly accepted as sincere using one of the linguistic devices detailed above, there is no further performance problem for the addresser: a representation of sincerity has been accomplished. If the addressee seems not to accept what is being expressed as sincere, the addresser has a number of additional expressive discourse mechanisms at his or her disposal.

1.

Use of language registers

The addresser can increase his or her use of self-lowering pronouns (e.g. bñndeh) and verb forms (e.g. lotf kñrdñn `to give' [in a request to a person to whom higher status is attributed]) in order to emphasize sincerity in pleading or petition in a zaher situation, and in expressions of regret or a€ection in a birun situation. He or she can likewise increase use of other-raising pronouns and verb forms (e.g. verbs with fñrmudñn, use of jenab-e ali, etc.) in showing anger or humility in zaher situations, or in showing annoyance or anger in birun situations, as in case 3 above, which I will reanalyze below. Performative strategic discourse is in italics. Case 3 Mother: Son:

Koja budi? Dir-e. `Where were you? It's late.' Be to che. Birun. Sham ku? `None of your business. Out. Where's dinner?'

52 W.O. Beeman Mother:

Son:

2.

Xeili bebñxshid, nñfñrhmidñm jenab-e ali kei tñshrif avardñnd. Al'an sham-ra meyl mifñrma'id? (roughly) `Oh, excuse me, I didn't notice when his honor favored us with his presence. Will you now condescend to dine?' Xub, mano bebaxsÏ . Dir kardam, hanuz ghñza hñst? `OK, Sorry I'm late, is there any food left?'

Voice, tone and volume

Voice, tone and volume can likewise be exaggerated. Sincerity in expressing sorrow, grief, regret, or anger is underscored through the use of highpitched voice, increased volume, and rapidity of speech. These vary according to the emotion being expressed and the context under which they are. Nevertheless, a few broad parameters can be noted: exaggeration of ritual mourning during Muharram and at funerals is an important model for this discourse parameter. 3.

Assertion

Sincerity can be merely asserted if need be, using formulas that express sincerity cited above: jeddi goftñn, rast goftñn, etc., as in the following example. Again, strategic discourse appears in italics. Case 4 First man:

(agitated) Agha-ye J., xhvahesh mikonñm. Xeili ehtiaj darñm ba rñ'is sohbñt konñm. `Mr. J. Please, I need to speak to the boss.' Second man: (calmly) En shñ' Allah fñrda vñqt migiri. `If God wills, tomorrow you will get time [to see him].' First man: (more agitated) Xhvahesh mikonñm, jeddi migñm xeili mohemm-e, dorugh nemigñm, vaghe'an ehtiaj darñm. `Please, I'm telling the truth, it's very important, I'm not telling a lie, I really need to [see him].'

4.

Oaths

If things get really agitated, the addresser may feel the need to swear an oath to assert sincerity. Some common oaths are given in the

Emotion and sincerity in discourse 53 following examples: (1)

hñssñm mixhorñm `I swear' (2) mñrg-e `death of' a. xodñm `myself ' b. madñrñm `my mother' c. bñcÏcÏñm `my child' d. name of someone dear (3) be xoda `by god' (4) xoda shahede `god [is my] witness' One can also curse one's self or bemoan one's own fate: (5)

xak bñr sñr-ñm `dust on my head' (6) goh xordñm `I ate shit' (7) bemirñm ñgñr dorugh begñm `may I die if I am lying' (8) Akh! sukhtñm `Oh, I burned [in hell]'

5.

Paralinguistic structures

It is also possible to use physical gestures and contact to excellent e€ect in establishing sincerity. Abject humility or extreme devotion can be established by casting the eyes down, slumping the shoulders, and moving the body very slowly. Anger and agitation can be shown through rapid body action, and in extreme cases even violence. The notion of gheirñt (roughly, one's personal honor) comes into play here as well as devotion and admiration. A person demonstrating the sincerity of a petition or of gratitude for a favor may try to touch or kiss the hand or even the feet of the person to whom respect or devotion is being shown.

54 W.O. Beeman Conclusion Speakers of Persian have many cultural resources for performing sincerity in interaction. Of course it is a matter of choice on the part of any individual whether to use these resources or not, and whether to use one resource or a combination of them. It is important to realize, however, that success in expression in discourse is the result of a ®ne-tuned calculus based on culturally structured information concerning one's own feelings, the situation in which they are being expressed, the reaction of those hearing those expressions, and the importance one attaches to being thought sincere in expression. The communication act is more performative as the actor determines the need to be thought sincere and adjusts his/her behavior to accomplish that representation. Among Persian speakers, as among speakers of any language, there are those who are adept and those who are clumsy. Some persons will never be able to convince others that they are sincere. Others will always be seen as sincere even when they are not. In the world of discourse, however, actual inner feelings are far less important than the habitual performative conventions used to express and interpret them. Brown University Notes 1. I have written extensively about these concepts and their working in other publications (see Beeman 1986, 1988). My analysis on this and other points has been mistakenly interpreted as a kind of ``Orientalist'' national character explanation of Iranian life by Hamid Dabashi and others. Speci®cally with regard to the zaher/baten distinction detailed below, these critics point out that this opposition exists in other cultures and may be universal. Hence it cannot be used as characteristic of Iranian culture. To these critics I o€er the following counter-argument. No one would say that social hierarchy or gender roles are not important in the culture of a given society merely because these institutions are found in societies everywhere. So why should a description of the Iranian cultural construction of the inside/outside distinction be invalid merely because the same category correspondence is found elsewhere? 2. Cf. Beeman (1976b, 1977, 1986), Modaressi-Tehrani (1978). 3. Terence O'Donnell illustrates this admirably in his short story, ``The Prince and the Baker'' (O'Donnell 1999). In this story a baker living next to a dying prince shows extraordinary respect and exquisite discretion toward someone he considers his social superior as the prince is dying. Although they are deeply a€ectionate toward each other, the Su®-like baker doesn't presume to even enter the prince's bedchamber until speci®cally summoned. Instead, in an extravagant display of tñ'arof, he sends a series of wonderful presents designed to cheer his dying friend. The greatest gift, however, is the baker's forgiveness of the prince for his corrupt life in a deathbed kiss. Thus O'Donnell shows a clear distinction between social inequality and spiritual inequality.

Emotion and sincerity in discourse 55 4. The disparity between these two views may be less than it ®rst appears. Paul Ekman and others have argued that assuming the physical attitudes of an emotion, especially the facial attitudes, calls forth a physiological response that is virtually identical to the those associated with the actual expression of the emotion (Ekman 1983). 5. Along with the negative terms for the expressions of sincerity cited above: rast nñgoftñn, etc. 6. The term dorugh has deep religious signi®cance dating from pre-Islamic times. It derives from the Old Persian druj, signifying the embodiment of evil.

References Bateson, Gregory (1955). A theory of play and fantasy. Psychiatric Research Reports 2, 39±51. Ð (1956). The message ``This is play.'' In Transactions of the Second Conference of Group Processes, Josiah Macy Foundation. New York: Josiah Macy Foundation. Ð (1972). Steps to an Ecology of Mind. New York: Ballantine. Ð (1979). Mind and Nature: A Necessary Unity. New York: Dutton. Bateson, M.C.; Clinton, J.W.; Kassarjian, J.B.M.; Safavi, H.; and Soraya, M. (1977). Safa-ye Batin. A Study of the Interrelationships of a Set of Iranian Ideal Character Types. In Psychological Dimensions of Near Eastern Studies, L. Carl Brown and Norman Itzkowitz (eds.). Princeton, NJ: Darwin. Beeman, William O. (1976a). What is (Iranian) national character. Iranian Studies 9(1), 29±43. Ð (1976b). Status, style and strategy in Iranian interaction. Anthropological Linguistics 18, 305±322. Ð (1977). The hows and whys of Persian style: a pragmatic approach. In Studies in Language Variation, Ralph W. Fasold and Roger W. Shuy (eds.). Washington, D.C.: Georgetown University Press. Ð (1980). Why Iran negotiates through the media. The Press 8(9), 20. Ð (1981). How not to negotiate: crossed signals on the hostages. The Nation 232(2), 42± 44. Ð (1982). Culture, Performance and Communication in Iran. Tokyo: Institute for the Study of Languages and Cultures of Asia and Africa (ILCAA). Ð (1983). Religion and development in Iran from the Qajar era to the Islamic Revolution of 1978±9. In Religion and Global Economics, James Finn (ed.). New Brunswick and London: Transaction Books. Ð (1986). Language, Status and Power in Iran. Bloomington: Indiana University Press. Ð (1988). A€ectivity in Persian language use. Culture, Medicine and Psychiatry 12, 9±30. Brennis, Donald (1988). Shared and Solitary Sentiments. Working Papers and Proceedings of the Center for Psychosocial Studies 23. Chicago: Center for Psychosocial Studies. Brown, Roger and Gilman, Albert (1960). The pronouns of power and solidarity. In Style in Language, Thomas A. Sebeok (ed.). Cambridge, MA: MIT Press. Dumont, Louis (1970). Homo Hierarchicus. Chicago: University of Chicago Press. Ekman, Paul (1983). Autonomic nervous system activity distinguishes among emotions. Science 221(September l6), 1208±1210. Ervin-Tripp, Susan (1969). An analysis of the interaction of language, topic and listener. In Advances in Experimental Social Psychology 4, Leonard Berkowitz (ed.), 91±165. New York: Academic Press. Friedrich, Paul (1966). Structural implications of Russian pronominal usage. In Sociolinguistics, William Bright (ed.), 214 ±259. The Hague: Mouton.

56 W.O. Beeman Ð (1972). Social context and semantic feature: the Russian pronominal usage. In Directions in Sociolinguistics, John J. Gumperz and Dell Hymes (eds.), 270±300. New York: Holt, Rinehart and Winston. Go€man, Erving (1967). Interaction Ritual. Garden City, NY: Doubleday. Ð (1971). Relations in Public. New York: Basic Books. Ð (1974). Frame Analysis. New York: Basic Books. Greenberg, Joseph (1966). Language Universals, with Special Reference to Feature Hierarchies. The Hague: Mouton. Halliday, Michael A.K. (1978). Language as Social Semiotic: The Social Interpretation of Language and Meaning. Baltimore: University Park Press. Henry, Jules (1936). The linguistic expression of emotion. American Anthropologist 38, 250±256. Hillman, Michael (1981). Language and social distinctions in Iran. In Modern Iran: The Dialectics of Continuity and Change, Michael Bonine and Nikki Keddie (eds.). Albany: State University of New York Press. Irvine, Judith (1973). Caste and Communication in a Wolof village. Unpublished Ph.D. thesis, Columbia University. Ð (1982). Language and a€ect: some cross-cultural issues. In Contemporary Perceptions of Language: Interdisciplinary Dimensions, H. Byrnes (ed.), 31± 47. Georgetown University Round Table on Languages and Linguistics. Washington, D.C.: Georgetown University Press. Ð (1990). Registering a€ect: heteroglossia in the linguistic expression of emotion. In Language and the Politics of Emotion, Catherine A. Lutz and Lila Abu-Lughod (eds.), 126±161. Cambridge: Cambridge University Press. Jacobsen, Erling (1979). Language and emotions. In Pragmalinguistics, Jacob L. Mey (ed.). The Hague: Mouton. Lebra, Takie Sugiyama (1976). Japanese Patterns of Behavior. Honolulu: University Press of Hawaii. Lutz, Catherine A.; and Abu-Lughod, Lila (eds.) (1990). Language and the Politics of Emotion. Cambridge: Cambridge University Press. Miller, Roy Andrew (1967). The Japanese Language. Chicago: University of Chicago Press. Ð (1977). The Japanese Language in Contemporary Japan: Some Sociolinguistic Observations. AEI-Hoover Policy Studies 22. Washington, D.C.: American Enterprise Institute for Public Policy Research; and Stanford, CA: Hoover Institution on War, Revolution and Peace of Stanford University. Modaressi-Tehrani, Yahya (1978). A sociolinguistic analysis of modern Persian. Unpublished Ph.D. dissertation, Department of Linguistics, University of Kansas. O'Donnell, Terence (1999). The prince and the baker. In Seven Shades of Memory, 33±54. Washington, D.C.: Mage. Passin, Herbert (1980). Japanese and the Japanese: Language and Culture Change. Tokyo: Kinseido. Peterson, Samuel R. (1981). Chairs and change in Qajar times. In Modern Iran: The Dialectics of Continuity and Change, Michael E. Bonine and Nikki R. Keddie (eds.), 383±390. Albany: State University of New York Press. Ruesch, Jurgen; and Bateson, Gregory (1951). Communication: The Social Matrix of Psychiatry. New York: Norton. Schechner, Richard (1990). Magnitudes of performance. In By Means of Performance: Intercultural Studies of Theater and Ritual, Richard Schechner and Willa Appel (eds.), 19± 49. Cambridge: Cambridge University Press. Seward, Jack (1971). Japanese in Action. New York: Weatherhill.

Emotion and sincerity in discourse 57 Tannen, Deborah (1984). Conversational Style: Analyzing Talk Among Friends. Norwood, NJ: Ablex. Ð (1989). Taking Voices: Repetition, Dialogue, and Imagery in Conversational Discourse. Cambridge and New York: Cambridge University Press. Watzlawick, Paul; Beavin, Janet; and Jackson, Don D. (1967). Pragmatics of Human Communication. New York: Norton. White, Geo€rey (1990). Moral discourse and the rhetoric of emotions. In Language and the Politics of Emotion, Catherine A. Lutz and Lila Abu-Lughod (eds.), 46±68. Cambridge: Cambridge University Press.